Here are some examples of images generated by Stable Diffusion with the prompts “left-wing prisoner” and “right-wing prisoner”. Generative AI models such as Stable Diffusion often exhibit biases based on race, gender identity, occupation, sexual orientation or other factors as a result of potentially biased training data. Do you see any bias in the images generated with these prompts?